Understanding the relationship between neural activity and cerebral blood flow is critical for interpreting hemodynamic signals, such as those measured with fMRI. It has long been assumed that blood flow to a brain region reported the average, or linear summation, of local neural activity. Recent work has cast this simplistic model into doubt. This proposal will use in vivo two-photon imaging, in close coordination with computational analysis methods, to distinguish between two alternative hypotheses of how neural activity is coupled to changes in blood flow. In one model, a 'democracy', blood flow is controlled by a linear sum of all neural activity. Alternatively, in an 'oligarchy', small groups o highly active neurons exert a disproportionate amount of control over blood flow, resulting in non-linear neurovascular coupling. Computational modeling will be used to test if the observed linear or non-linear coupling can be mechanistically explained by the production and diffusion of nitric oxide (NO). The proposed experiments will be performed in the olfactory bulb of rats, where discrete subpopulations of neurons (glomeruli) will be visualized and stimulated with odors. Two-photon microscopy will be used to simultaneously measure neural activity and blood flow in defined neural populations and single blood vessels. Targeted applications of drugs will be made to increase or decrease the neural activity in a single glomerulus. These experiments will be guided by real-time data analysis to determine the optimal stimulus or pharmacological perturbation in order to obtain a more accurate quantification of the linearity or nonlinearity of neurovascular coupling. In parallel, computational models will be constructed to test if the generation and diffusion of NO, a potent vasodilator, can account for the observed neurovascular coupling. This proposal is a collaboration between the labs of Dr. Serge Charpak, who has expertise using two-photon microscopy to simultaneously measure neural activity and blood flow changes in the olfactory bulb, and that of Dr. Patrick Drew, who has a background in computational neuroscience and has developed novel hemodynamic data analysis methods. The combination of these two approaches will yield a quantitative understanding of how blood flow changes relate to neural activity, and a determination of the mechanisms underlying neurovascular coupling. Hemodynamic signals, such as those measured by fMRI, are extensively used in inferring brain activity non-invasively, and being able to convert these hemodynamic signals into neural activity would be invaluable in diagnosing cognitive and neurological disorders. However, what specifically these changes in blood flow tell us about neural activity is not known. This proposal will result in a quantitative understanding of how neural activity is translated into hemodynamic signals, which will have immediate application to the interpretation of human imaging studies. This proposal will support undergraduates in mentored summer research projects, building on Dr. Drew's track record of mentoring women and underrepresented minorities in undergraduate research. The results will be incorporated into an interdisciplinary undergraduate class taught by Dr. Drew, Physical principles of living organisms, which applies physics and engineering principles to the study of biological systems.